Can artificial intelligence save the NHS?

According to the Office for Budget Responsibility, the NHS budget will need to increase by £88billion over the next 50 years if it is to keep pace with the rising demand for healthcare in the UK. But with the 2017 Budget showcasing a massive leaning towards building up its Brexit reserves and allocating a mere £100 million for 100 onsite GP treatment centres in A&Es across England, the NHS is justifiably bracing itself for a painful future. 

With £20billion worth of cuts scheduled by 2020, combined with fierce warnings that the UK’s health services are on the edge of an unprecedented crisis, the urgent call for solutions to be brought to the healthcare table has incontrovertibly intensified.

With deep cuts looming, it’s time to properly consider how Artificial Intelligence can answer this call and shed light on how its technologies could provide the healthcare industry with some much-needed respite and real solutions to meet the ever spiralling rise in demand for healthcare. 

Crunching the data

The issue of voluminous data that draws relentlessly on healthcare professionals’ resources is something that could benefit significantly from the implementation of an AI-based system.

It has been estimated that it would take at least 160 hours of reading a week just to keep up with new medical knowledge as it's published, let alone consider its relevance. It soon becomes apparent then, that it would be physically impossible for a doctor to be able to process all of the patient information as well as digest insight from new materials and medical journals, and still be able to treat patients. 

Imagine a scenario wherein supercomputers could process the information – and far more efficiently, too – making sense of the sheer quantity of data, flagging any relevant information to the doctors and nurses that might be pertinent to a patient’s case, and providing them with access to up-to-the-minute and highly applicable insight in the field.

Such an AI system would effectively unshackle medical professionals from these time-consuming processes, freeing them up to focus on work that requires human skills. Contrary to popular belief that AI will result in mass job losses, the implementation of AI systems in this instance would actually augment the roles and skills of the human workers – performing the tasks they don’t have the time or capacity to do. Moreover, this rapid analysis and provision of data would enhance the overall efficiency of the human decision-making processes. And so, rather than replace jobs, the AI systems would empower human services.

Case in point

This is exactly what IBM Watson has been working on in collaboration with Memorial Sloan-Kettering Cancer Center. World-renowned oncologists have been training Watson to compare a patient’s medical information against a vast array of treatment guidelines and research to provide recommendations to physicians on a patient-by-patient basis.

Supporting evidence is provided for each recommendation in order to provide transparency and to aid in the doctor’s decision-making process, and Watson will update its suggestions as new data is added. Watson is being used to facilitate access to the best of oncology’s collective knowledge, therefore demonstrating how this can be applied across the entire medical profession.

Having recognised the potential that AI tech can bring to the wider industry, community healthcare service Fluid Motion has rolled out pilot trials in a bid to overcome the challenges they face in relation to cost, staffing, efficient decision-making processes and data crunching.

Born from the frustration of facing barriers presented by the current healthcare system, Fluid Motion’s group aquatic therapy programme is a tailored rehabilitation concept that has been designed to be both fun and beneficial for people with a range of musculoskeletal conditions, with an overall aim to treat, manage and prevent such conditions. 

With one in five GP appointments being related to musculoskeletal disorders – translating into a cost to the UK economy of £24.8 billion per year due to sick leave – the need for fast and effective healthcare solutions is clear. But the challenge, as indicated Ben Wilkins of Fluid Motion, is that while these programmes are successful, there simply aren’t enough professionals to sustain the growing levels of demand for the service. Additionally the very nature of the programmes means that they depend heavily on vast amounts of data input and analysis to determine the right solution.  

Fluid Motion recognised that, if they could generate these rehabilitation plans automatically, it would allow them to lower their staff costs and increasing their reach. Fitness Instructors could quickly generate a high-quality tailored plan based on a model of the Physiotherapist and Osteopath’s expertise, modelled in AI-powered cognitive reasoning platform, Rainbird.

Rainbird modelled the knowledge of Fluid Motion’s qualified physiotherapists and osteopaths, including the suitability of numerous exercises to individual patient symptoms, and added it to an interface that could be accessed by Fluid Motion’s network of fitness instructors. The tool allowed them to create a tailored, illustrated rehabilitation plan for patients, based on the results of an initial interaction with a virtual physiotherapist or osteopath. 

The next step will be to provide access to patients directly so that they can create their own rehabilitation plans. Patients will have the facility to give feedback so that Rainbird can learn and, where necessary, adapt their plan or make alternative recommendations if specific exercises are uncomfortable.

Fluid Motion has since been able to track and reflect on participants’ progress in real-time, meaning the data can be utilised to improve clinical decision-making in rehabilitative healthcare.  The application of AI helps patients get better sooner, and prevents pain and disability for longer.

The time and cost saving possibilities resulting from the implementation of such a programme are indubitable. According to Wilkins, the cumulative cost for a healthcare professional per session is £75 (£50 for hiring an Osteo/Physio for the whole session and £25 to pay them to review feedback data to make recommendations). Fluid Motion sessions now only cost the company £35 (for a Fluid Motion fitness instructor) and £25 (for pool hire), making previous session delivery 125% more expensive. With this model, it means that Fluid Motion can charge participants less than the average price of a swim to attend sessions.  

Up to this point, Fluid Motion had been subsidising cost with grant payments, but now the company breaks even each session. Moreover, this is a model which is scalable. As a result of this initiative, Fluid Motion is now working to become an organisation that provides support and treatment for musculoskeletal health conditions alongside the NHS. 

Indeed, the Fluid Motion case study clearly illustrates how challenges in healthcare can be overcome through the implementation of AI systems, and also highlights the potential time and cost saving benefits that the NHS could reap, if such an approach were adopted.

By mapping knowledge of some of the medical roles that are in high demand, there are many ways that the technology can help to streamline some of the more rudimentary elements of those roles.  This would free up time to devote to face-to-face consultancy that would have the most impact for patients, reduce waiting times and even enable medical professionals to engage in a more personalised service.

This application of AI has the potential to address the rise in demand for NHS services, whilst ensuring that doctors and nurses spend more time doing the work that they are trained to do; treating patients to the best of their ability. Indeed, with the assistance of AI-powered technologies, the NHS may not only survive the crisis but, like the Phoenix, rise from the ashes to achieve its original goal of ‘bringing good healthcare to all’.

Katie Gibbs, Head of Accelerated Consulting, Aigen
Image Credit: John Williams RUS / Shutterstock